Forecasting Natural Gas Futures Price Volatility of the United States: National Versus State‐Level Climate Concern Indexes
使用GARCH-MIDAS模型，结合全国和州级气候关注指数预测美国天然气期货波动，发现州级数据（尤其最小二乘加权组合）优于全国指数和基准模型，带来显著统计和经济收益。
ABSTRACT This paper uses GARCH‐MIDAS to predict US natural gas futures volatility using national and state‐level Climate Concern Indexes (CCIs). We find that both national and state‐level CCIs positively affect price volatility. Notably, models using state‐level data—specifically those utilizing least‐squares (LS) weighting combinations—surpass the GARCH‐MIDAS‐GECON benchmark and models relying solely on national CCI. These findings deliver substantial statistical and economic utility gains. Our results underscore the importance of incorporating heterogeneous climate concerns across US states to capture varied demand‐supply conditions when forecasting energy market volatility.